Mmmbeats

The high-contrast streaking issue seems to be rearing its head a bit unfortunately.

(not personal experience, just what I'm seeing online)

. 

1 hour ago, herein2020 said:

 

 

 

I have read that in a true lowlight situation that CLOG3 would be the better color profile, but I haven't shot at night yet with it so I have no idea. 

I did some proper low light shooting with the camera (a completely unlit tunnel deep underground, lit only by somebody's helmet torch). The results blew me away. I truly think c log 3 is redundant on this generation of cameras. 

Regards the straight adaptor, the simplest one (no drops-ins, etc.) is only £80. Probably worth getting just for the extra flexibility it offers. 

Yeah, I'm now very much in the @herein2020 was right camp.

I think that description of the effect of the speedbooster was spot on. 

From a practical point of view, I'm not sure that the speedbooster actually needs to be held in place by the screws. I suspect it may be optional. 

I run a dual system strategy on my GH5S (native lenses and adapted lenses jumping on and off during the day) and it works absolutely fine. The whole thing is second nature to me now. 

On 2/22/2022 at 1:25 AM, herein2020 said:

 

Paris France? Paris is legendary in the fashion world, along with Milan, NY, and Miami. I am in Tampa FL, so the shows here are nowhere near the Paris league. Even Miami is a massive difference, I shot the Miami Swim week show last year and the lighting was picture perfect. Here in Tampa FL, I've literally had to light the models with not much more than an on camera light.

Since COVID hit, it has been even worse because most of them are held outdoors now and outdoors at night there's no guarantee of a power source to power my bigger lights.

That video looked great, I have the 50mm F1.4 so with the speedbooster and a little fill light I should be able to make it work. I will definitely do everything I can to stick to 800ISO, but when you are shooting wide open sometimes your only choice is to start cranking the ISO if it is still too dark. I've even had to drop my shutter speed to 1:1 in the past just to get more light. The increased motion blur was worth the alternative.

 

I remember that discussion, personally I'd just be happy with 32bit float audio, but it wouldn't really help me in many situations. When using wireless lavs/mics the audio can clip at the transmitter or receiver before it even reaches the camera so even with 32bit float audio, it wouldn't help my typical audio scenario.

My main problem is there's not that many stabilized EF lenses, so I will be stuck with no stabilization or digital stabilization for lenses like the 50mm F1.4.  I don't know how good Sony's stabilization is, but I know my GoPro's stabilization is incredible as well. The latest GoPro also performs horizon leveling which mine does not have. The C70's stabilization on the other hand is nowhere near as good.

I noticed that as well, I never shot ALL-I with the GH5 but when I tried it, the cards failed almost immediately. I do think the C70 will be fine at 4K30FPS which is only 160Mb/s, 4K60FPS is 260Mb/s and it is the main framerate I use when handholding or using a gimbal, so I will need to do a lot of testing with that one. Longform talking heads stuff is all 30FPS anyway, so I already ordered the 1TB cards.

 

Everything I've read says don't ETTR with the C70. Just shoot down the middle and it will be fine. For lowlight use CLOG3, for everything else use CLOG2.

The DGO sensor really takes care of noise spectacularly well.  From what I have seen, read and experienced so far there's really no need to use C Log 3 at all with this camera. 

18 minutes ago, Django said:

actually @herein2020 is correct: 

a speed booster as its name indicates .. boosts the aperture speed. you gain a stop of light. but DoF remains the same. that's where sometimes the confusion comes in.

it is until it hits the speed booster. a speed booster adds an extra glass element changing the optical design of the original lens. It demagnifies and hence concentrates light adding a stop when it hits the sensor. this article explains it better using the same 50mm example too:

I've already explained above that the entrance pupil does not change. However, given that we've now reduced the focal length (from 50 to 35), the formula for calculating f-number has changed. The entrance pupil is still 25mm in diameter, so it now gives us 35/25=1.4. So a 50mm f/2 lens, speedboosted, becomes (not effectively, but actually) a 35mm f/1.4. That's how Speedboosters 'magically' create an extra stop of light. To clarify, our 35mm f/1.4 will give you an exposure one stop brighter than the original 50mm f/2, whether that 50mm is used on full frame or S35.

https://www.alex-stone.com/2020/10/29/busting-speedbooster-myths/

 

I was just looking at the same resource to check my thinking on this. 

I'll give it another look tonight. 

(unless I'm too busy playing with my new C70, which I'm literally on the way to collect!). 

7 hours ago, M_Williams said:

He's actually not correct at all. The speedbooster doesn't make it a 50 f/1 lens as he states it. It makes it (roughly) a 35mm f/1 lens. Which on APS-C gives you... a 50mm FOV with the DOF of a 50/1.4. As I said, you cannot get the extra stop of light without shallower DOF, or vice versa. A speedbooster can't make the DOF shallower than the lens would be on FF, which seems to be his issue with it. And the lens is still transmitting the same TOTAL light as a 50/1.4, and the sensor captures the same TOTAL light as a full-frame sensor, hence the DOF is the same.

If it made the lens a 50 f/1.0 then it would have the same FOV with and without the speedbooster.. obviously we know speedboosters widen the FOV. From the same distance, a 35mm f/1.0 will have about the same DOF and FOV as a 50/1.4 on FF.

I'm coming round more to your (field) of view now.  I haven't gone back over all your statements, but I do realise I was fundamentally misstating the role of focal reducers in the imaging chain, so I'll take another look at this.

That's twice this week I've been wrong about something.  Disconcerting, as I am usually only wrong about something around once a decade.  I had an exhilarating run from my 13th birthday till turning 30 when I was right about *absolutely everything* 😅. 

7 minutes ago, Django said:

Also pretty annoying DR boost mode turns base ISO to 2000. I'd rather it be at a low value like 400.

At 2000 you're going to need ND's or close the iris during daylight which is when you want that DR boost.. odd choice.

I doubt it was a choice.  Likely something technical to do with how they stack the exposures.  I doubt Panasonic were too pleased about that aspect of it either.

16 hours ago, herein2020 said:

With the S5 I typically exposed a bit to the right and to me anytime to have to do that it is a bit of a guessing game, with the C70 you just property expose so to me that is easier especially when you have time to check the false color.

Exposure is very much a personal thing, but I do now find that the built in exposure meter is probably the worst way to do it. 

V Log and C Log 2 are so similar I'm hoping I'll be able to use very similar, or even identical skintone targets with both (I'll be using C70 and likely a GH6 as B-cam). 

On my GH5S the inbuilt meter does a surprisingly good job at making exposure for linear profiles but it's all over the place when it comes to V Log. 

The exposure targets I use are:

Darkest skintone highlights : 42 IRE

Lightest skintone highlights : 55 IRE

Of course you still have to make adjustment decisions once these targets have been considered, but it's a system that has worked very well for me up to now. 

No grey card, no meter, no trouble with noise. Just zebras or false colour required. 

I'm expecting the targets for C Log 2 to be very similar. I'm going to do some tests this week. 

16 hours ago, herein2020 said:

 think that the C70 and CLOG2 is easier to grade right out of camera and easier to expose than VLOG. I just set the exposure for the WFM so that the highlights touch 80 for outdoors shoots. If the sun is in the frame I let that peak hit 100. 

V log and C log 2 are very similar. I mainly use target skintone IRE's to fix exposure and the figures (which I can't recall from the top of my head) are very similar for each. 

I've found V Log (and V Log L) very easy to work with both in exposure and grade to be honest. 

In other news - I just purchased a C70, which I will be collecting on Monday 😍.

On 2/27/2022 at 3:38 PM, M_Williams said:

I'm not sure what you mean by "DOF remains the original DOF." It would be the same as if you mounted the lens on a full-frame sensor camera, but the DOF would be shallower with the speedbooster than if you used a regular adapter on the C70's sensor. The 1-stop extra light and shallower DOF go hand-in-hand; can't have one without the other.

On 2/28/2022 at 1:44 AM, herein2020 said:

You are incorrect, the speed booster adds a stop of light so it turns a 50mm F1.4 into a 50mm F1.0, however, the DOF is still the same as an F1.4, vs a true F1.0.

You are both correct, as this is just a different way of describing the apparent DoF.

@herein2020 is technically correct (yes, yes everybody - 'the best kind of correct' 😉), and @M_Williams is correct for practical purposes.

DoF is in fact the same whether using the speedbooster or not.  However, in order to achieve the same framing on the 50mm lens example you would have to alter the distance between the camera and the subject, thus altering the DoF.

I much prefer herein2020's way of stating it, but I have come to acknowledge that endlessly explaining this introduces just as much confusion as it clears up. 

5 hours ago, kye said:

 

That makes sense.  Potentially then it's quite a different equation for your ND setup, but I keep thinking that all you need to do is work out what the maximum and minimum amount of ND you will need in daylight and just buy a fixed ND that compliments your vND to give you that range.  Then after sunset you remove the fixed ND and your vND should cover the range of shooting under artificial lights.  ISO 2000 should be a good base ISO for most low-light situations, so you should be using your vND even after dark anyway.

I think this is what most people are going to do.  It troubles me though because you are shooting through 3 bits of tinted glass (2 polarised planes in the variable ND).

My experience of variable ND (Tiffen 2 - 8 stops) is mixed.  Around 85% of the time it is fantastic with little to no noticeable cast.  Many shots I forget that there was even any ND used.

But the times when problems are introduced can be very hard to predict.  Weird casts seem to appear based on such criteria as window coatings in buildings (especially office blocks and new-build public buildings), light bouncing off buildings, the sun at certain angles to the filter, etc.   The cast can also range from muddy brown to grey-blue (at high ND's).

It can also play funny with other filtration, such as diffusion.  All of which makes me a little nervous about stacking variable ND with more stuff. 

On 3/1/2022 at 4:29 AM, kye said:

In terms of ISO2000 compared to 800 (wasn't 800 the native ISO for Vlog on the GH5?) isn't it only slightly more than a stop more exposed?  If so, you only need to add a single-stop of ND to your setup and it's the same.  Or am I missing something?

This is a decent point actually, but the GH5 native ISO for V-log is *optimal* rather than compulsory.  You can expose below it and take a slight DR hit.  Whereas the GH6 ISO 2000 requirement is a *minimum* for DR Boost, and if you want to expose at a lower gain setting you take a huge hit by switching to non-boost.

16 hours ago, mercer said:

I guess that depends...

If the GH6 didn't have ProRes, I would yawn at this release. But I acknowledge that some of the updates could be useful to some people.

But without a doubt, the ProRes propels the GH6 into next level territory... especially with the updated IBIS.

I'm hoping the future ProRes updates will offer the 48p/60p in 1080p/4K but since I don't really shoot too much slow motion, it's not that big of a deal.

The bump in dynamic range is minimal... what... about a stop and it's still under 12 stops. I'll take any extra they want to give, but that's not so great to go crazy over.

Actually, I know there are limitations to what m4/3 can do, but how did BM get 13 stops on the OG Pocket, 9 years ago, but Panasonic can't seem to break 11.5 stops, even with the Go-Go-Gadget-Boost mode and full vLog?

I'm the complete opposite.  I wasn't really that excited for the release and had no real intention of updating. 

I assumed that Panny would go for some kind of headline-grabbing big feature, but instead they seem to have really gotten to the DNA of what makes the line so appealing - great useability, powerful processing, features for real-world filmmaking.  I have to say I'm impressed - and sold!

A stop of dynamic range is actually a huge deal.  If you're trying to bring a window into some kind of exposure range, that's half the amount of light you have to deal with.  That's like pulling a couple of big fixtures onto set.

There's really no point comparing the GHx to the BMPCC in my opinion because they are completely different tools.  No way you can engineer what BM made with full IBIS, mechanical shutter, weather sealing, etc.  They are just great at doing very different things.

35 minutes ago, mercer said:

Obviously this conversation was born from the GH6 release and its inclusion of ProRes. Well, I think it's quite simple. If you have no plans to use ProRes, there is no real need to upgrade from a GH5.

Not sure if you're being serious?

4K 60p 10-bit (the thing I most lament on an almost daily basis) 
Increased dynamic range (the GH5's biggest weakness)
High resolution mode stills (for landscapes, architectural, etc.)
4 channel audio recording....

1 hour ago, kye said:

I did consider that but didn't think that you'd be doing anything as radical as a green screen or background replacement etc.

What kinds of secondary adjustments would you really notice a poor quality codec?  Genuinely curious.

It's all about the quality of the isolation.  it can be tough to get, say, a clean skintone isolation even with a really strong codec at times.  If the colour data is thin you either end up with areas of skin that just won't key, or parts of the background that just wont separate.  I've had to just abandon secondaries at times.

Totally agree about Canon.  They seem to have some compression fairy dust.  The C200 8-bit was remarkable.

5 hours ago, kye said:

Interestingly, I thought for the test I would encode and re-encode each codec over and over again until one of them started breaking down, but I setup ffmpeg to do that on a 1080p version (to save processing time) and after 8 generations on h264 IPB at the same bitrate as Prores HQ I had to pause because the file looked the same.  

 

That's not good enough for me.  An 8-bit file from the same camera will generally 'look the same' (absent any banding) as a 10-bit file from the same source.  That doesn't mean its going to perform as well in post.

5 hours ago, kye said:

Of course, if you're doing green-screen or other hard-edged keying, cropping in post, stabilising significantly, compositing and other VFX operations then that's a different story, but those are specialist applications and not general ones.

 

I think what people miss out here is secondary grading isolation, which also benefits from a rich data environment. 

I don't consider this specialised at all.  Its a part of a well-rounded grading workflow.  I don't end up using it for every single project, but if anybody is not using secondaries at all, well... they probably should be in my opinion.

The main advantages of ProRes over H.264 and H.265 are smoother playback and better facility for bouncing down to further generations.  Its designed to work really well in a post production environment, which it does superbly in my experience.

Its no surprise at all that the codecs look similar at the same data rates, in fact it is H.265 that you would expect to look the best as it is the most efficient encoder.

Of course H.264 looks good at reasonable data rates.  If it didn't it wouldn't have been utilised as an acquisition codec by just about every manufacturer under the sun at some point.

The point is that IPB codecs are horrendous in a post environment.  People are excited by ProRes 422HQ in the GH6 because, unlike the other two codecs, it is *guaranteed* to have high data rates (1.9 Gbps is what they are saying), and it will fit straight into our post workflows.

1 hour ago, projectwoofer said:

Yeah, I quickly tested that. It is much less pronounced with HLG rather than V-Log, which is not a good choice for underexposured images anyway and only starting to be visible with 5 or more stops of underexposure. In which point the image lacks so many details and contrast that is in my humble opinion unusable anyway even without the horizontal lines. When and why would anyone want to use an image underexposed by 5 or 6 stops is beyond me to tell you the truth. Not in my use scenario...thanks for pointing this out though, it was a fun experiment!

Over / under tests are not just literally about over and underexposing, they also give you information about what is happening at the various extremes of your normally exposed image.

For example, if a camera performs badly in over tests with colour degradation, you can reasonably surmise it is going to have difficulty retaining highlight information in a properly exposed shot. 

11 hours ago, hyalinejim said:

I noticed that as well. 

1 hour ago, hyalinejim said:

As is normal exposure in daylight! I don't really fancy the idea of stacking NDs just to get an exposure (my variable ND is 6 stop maximum and I calculate that you'd need 10 stops ND to shoot f2.8 1/50s ISO 2000 in bright daylight). Aside from reflections etc it makes changing lenses a pain.

A three or four stop internal ND would be so useful in this case 

I have to admit I lose track of how much ND I'm actually using with my (up to 8 stops) variable ND.  It's around 4 stops back down to ISO 125 from ISO 2000.  How much more would you say is required to protect highlights in the scenario you describe?  A further 6 stops?  I'm finding it hard to be sure.

2 minutes ago, MrSMW said:

Shooting video at f22 is gonna be a thing.

Yup, that shallow DOF is sooooooo 2021.

Yes, perhaps this camera will set a trend!  Personally I like to choose my f-stop, but still...

6 minutes ago, hyalinejim said:

As is normal exposure in daylight! I don't really fancy the idea of stacking NDs just to get an exposure (my variable ND is 6 stop maximum and I calculate that you'd need 10 stops ND to shoot f2.8 1/50s ISO 2000 in bright daylight). Aside from reflections etc it makes changing lenses a pain.

A three or four stop internal ND would be so useful in this case 

Well quite.  To use this camera I think you're going to need to buy a big stopper photographers ND filter because variable ND usually looks awful at high levels.